In lookout, I can plot values from a CSV file into a graph, but 3000 values is many. Is there a way to use hypertrend to scroll and view the values? This would be much easier to compare subtle values.

[smigman]

I have no problem displaying 4ma-20ma values onto a graph, from a CVS file. What I need to do, is find a way to put that information into a scrolling graph, so I can see detailed values. 3000 values is too many points to decipher any subtle changes on a stationary graph.

[Chad E]

Hello smigman,

I wouldn't recommend putting 3000 values onto any graph. I'm not sure what you mean by scrolling, but that is still an overwhelming number of values for a graph. I suggest that you use multiple graphs or limit to that which most needs to be seen.

Regards,

Chad Evans

[smigman]

I thank you for your reply! That's a good idea.

Here's the deal.

The customer, which is a waste water treatment plant, in a major city, has specified that sewage levels at various remotes sites be recorded every 15 minutes (via DATA logger). The specs state that they need to record one month's worth of DATA, onto the logger (not quite 3,000, records). Which includes being able to view the results from files (CSV) retrieved into LOOKOUT, in a GRAPH format.

I know this sounds demanding, but in the real world, the job was specified but the owner, approved by engineers, bidded by several contractors and awarded.

I haven't been able to meet the contract, although the specs are vague, I am possibly thinking that the graph can viewed in "chunks". For exampl
e, seven day segments. Or even one day segments. They are not clear on how to view the graph. They can jump to any given day and view the results!

Thanks again.

[Khalid]

Hi,

Why don't you log data to the built-in database, Citadel? You can do this in addition to logging it to the CSV file (if the CSV is a requirement from your client).

The advantage of doing this -- logging to Citadel -- is that you can use the HyperTrend. HyperTrend as you know has scrolling, value look-up with cursor, and locating of breaks/min/max, etc.

Not sure how you are logging data to CSV file. One option you have is to do the reverse. That is, log data to Citadel first, and then query this data to generate the CSV. But if you already have the CSV logging implemented, you may want to just stick to it. SQL queries into Citadel will need some tries before you get them all right.

Hope this helps,

Khalid

ps:
if you want to trend data that is already been logged to the CSV, you can still shove this data into Citadel using the Logger object. IMP: try this on a new test database (which you can throw away) because entering data with custom timestamps is not that straight-forward with the Logger object, and you don't want to mess up your main Citadel database.